The incidence of human non-Hodgkin's lymphoma (NHL) in the US has nearly doubled over the previous 30 years. Epidemiologic research has evaluated a variety of factors potentially associated with risk of NHL but results have been inconsistent and cannot explain the dramatic increase in incidence. Several epidemiologic studies in humans have suggested that exposure to household environmental tobacco smoke (ETS), lawn care herbicides, and pesticides such as those contained in flea control products may plausibly increase risk of NHL, but these relationships have not been well evaluated. Epidemiologic studies of companion animals offer a complementary means by which to identify risk factors for cancers of relevance to both animals and humans. Because canine malignant lymphoma (CML) has been established as a model for NHL and domestic dogs share their living environment with their human owners, a study of environmental exposures and CML provides a unique opportunity to evaluate factors that may potentially relate to NHL. The proposed case-control study will use hospital records to identify cases of CML and 2 control groups of dogs, one consisting of dogs with benign tumors and one consisting of dog with non-cancer chronic diseases, seen at the Foster Hospital at Tufts University between 2000 and 2005. Owners of CML cases and controls will be sent a questionnaire to assess household ETS, use of flea control products, lawn care herbicides and other factors prior to diagnosis. Multivariate logistic regression will then be used to evaluate the relationship between environmental exposures and risk of CML. In addition, because the methods used in epidemiologic studies of companion animals are not well established, the proposed project will also include a validation study of ETS exposure measurement, and evaluate if owner characteristics are associated with likelihood of returning a mailed questionnaire. Owners of 1000 dogs presenting to Foster Hospital over a two-month period will be asked to complete a questionnaire measuring demographic and environmental factors, and to provide a urine sample from their dog. Urine samples from approximately 130 dogs will be assayed for level of cotinine, a metabolite of nicotine well established as a biologic marker for ETS exposure, and cotinine levels will be compared with level of household smoking reported by questionnaire. One year later, a second questionnaire will be mailed to all participants, and characteristics of respondent and non-respondent owners will be compared. Any differences will then be taken into consideration in the calculation of relative risks using a novel form of sensitivity analysis to produce more accurate estimates of the association between environmental exposures and risk of CML. Because dogs and their human owners share a common environment within the household, results from this study will have direct relevance to the on-going effort to identify factors that may increase risk of human NHL.